The British Journal of Psychiatry (2016)
209, 385392. doi: 10.1192/bjp.bp.115.177261
Review article
Clozapine v. first- and second-generation
antipsychotics in treatment-refractory
schizophrenia: systematic review and meta-analysis
Dan Siskind, Lara McCartney, Romi Goldschlager and Steve Kisely
Background
Although clozapine is the gold standard for treatment-
refractory schizophrenia, meta-analyses of clozapine for this
condition are lacking.
Aims
We conducted a systematic review and meta-analysis of
clozapine treatment for people with treatment-refractory
schizophrenia.
Method
We searched the Cochrane Schizophrenia Groups trial
register, PubMed and EMBASE and hand-searched key
papers for randomised controlled trials of clozapine for
treatment-refractory schizophrenia.
Results
Twenty-one papers with 25 comparisons were included. The
number needed to treat was 9. Clozapine was superior for
The development of clozapine over 50 years ago was seen as a
major breakthrough in the treatment of schizophrenia. However,
initial enthusiasm for its use was tempered by the risks of adverse
drug reactions, notably a cluster of agranulocytosis cases detected
in Finland in the 1970s.1 This led to the drugs removal from the
market in many jurisdictions. In spite of this, continued evidence
of its effectiveness for psychotic symptoms led to its widespread
reintroduction in the 1990s.2,3 Clozapine is now generally reserved
for people with treatment-refractory schizophrenia, usually
defined as two failed trials of other antipsychotic medications of
adequate dose and duration.47 In spite of the value of clozapine
in these situations there is still a reluctance to prescribe it, with
the result that usage rates vary widely.8 This reluctance may mean
long delays before people with treatment-refractory disorder are
given clozapine, resulting in poorer outcomes and exposure to
potentially hazardous antipsychotic polypharmacy.9 One way to
address this reluctance is to demonstrate the particular value of
clozapine in treatment-refractory schizophrenia. However, most
Cochrane reviews comparing clozapine with other antipsychotic
medications have not focused on such patients. Furthermore,
the sole meta-analysis that specifically examined the effectiveness
of clozapine in this patient group included only comparisons with
first-generation antipsychotic medications.10 That review found
clozapine to be superior but also noted that the effect was greatest
in studies of shorter duration and studies funded by manu-
facturers of clozapine. Since that time there have been several
further randomised controlled trials (RCTs) in people with
treatment-refractory schizophrenia that, importantly, compared
clozapine with second-generation antipsychotics. These may be
more relevant to current practice. We therefore undertook a
systematic review and meta-analysis of RCTs comparing clozapine
with other antipsychotics for treatment-refractory schizophrenia.
Outcomes were grouped into short-term and long-term categories
for total, positive and negative psychotic symptoms, adverse
positive symptoms in both the short and long term. In the
short term only clozapine was superior for total and negative
symptoms, with higher response rates. Both funding source
and dosage affected results. Higher baseline psychosis
scores predicted better outcomes for clozapine in a
meta-regression.
Conclusions
Clozapine is superior for treatment-refractory disorder but
if there is no response by 6 months medications with lower
adverse reactions should be considered.
Declaration of interest
None.
Copyright and usage
B The Royal College of Psychiatrists 2016.
drug reactions, study withdrawals and response to treatment.
Additionally, we conducted sensitivity analyses for the effects of
pharmaceutical industry funding, types of antipsychotic controls
(first or second generation), dosage and initial psychosis score.
Method
The study was registered with the PROSPERO international
database of prospectively registered systematic reviews
(CRD42014013134).11 Preferred Reporting Items for Systematic
Reviews and Meta-Analyses (PRISMA) statement recommendations
were followed for background, search strategy, methods, results,
discussion and conclusions.12
Eligibility criteria
We included all RCTs that compared people with treatment-
refractory schizophrenia taking clozapine with those prescribed
a first- or second-generation antipsychotic. Outcomes of interest
included psychotic symptoms (total, positive and negative),
adverse drug reactions, study withdrawal and response to
treatment. Published data in all languages were included and
translated into English.
Search strategy
We searched PubMed, EMBASE and the Cochrane Schizophrenia
Groups Trials Register from inception to 6 February 2015. In the
case of PubMed we used the following terms: (randomized
controlled trial OR controlled clinical trial OR randomized OR
placebo OR clinical trials OR randomly OR trial) AND (clozapin*
OR clozaril OR zaponex OR denzapin* or clopine). Experts in the
field were contacted and asked about unpublished data.
385
Siskind et al
Study selection
Studies were included if they were randomised and double- or
rater-blinded. Diagnoses included schizophrenia, schizoaffective
disorder or schizophreniform disorder. Participants had to have
demonstrated a resistance to treatment as defined by a failure to
respond to at least one trial (and preferably two) of a first- or
second-generation antipsychotic of at least 6 weeks duration at
dosage equivalents greater than 600 mg chlorpromazine. Studies
were included if they compared clozapine with any other first-
or second-generation antipsychotic medication. Studies were
excluded if there was extensive crossover between the clozapine
and control groups. All identified studies were screened at the title
and abstract level by two authors (L.M. and R.G.). Studies that
met the inclusion criteria on title and abstract review, or that
could not be excluded on the basis of information provided in
the abstract, were reviewed at full-text level. Snowball searches
of key papers and the included studies reference lists were
conducted. Narrative and systematic reviews, posters, conference
abstracts, case reports, letters to editors and other articles that
did not meet the inclusion criteria were cross-referenced for
additional potential sources of RCTs. Attempts were made to
contact first authors of included studies in cases where
information was missing.
Data collection
Data extraction was conducted by two independent researchers
(L.M. and R.G.). All discrepancies during each stage of study
selection, data extraction and quality assessment were resolved
by re-checking source papers. Extracted data were validated by
D.S. Data analysis was conducted by two authors (D.S. and
S.K.). We extracted data on the following: study duration, setting,
diagnostic tool and type of interventions (e.g. control medication,
first- or second-generation antipsychotic, mean age and standard
deviation, number commenced in study arm, mean dose of
clozapine or control medication, and dose of clozapine or control
medication). Doses of clozapine and control medications were
converted to chlorpromazine and olanzapine equivalents.13,14
These were used in separate meta-analyses comparing dose
equivalents for clozapine and the control medication for each
study. Where data in studies were missing or unclear, attempts
were made to contact the studys corresponding author.
Outcomes
Where possible, end-points were measured from commencement
of intervention. Data from studies were divided into short term
(less than 3 months) and long term (3 months or more). These
time frames were selected after data were extracted, based on an
approach used in a previous meta-analysis.15 Where multiple
outcome time points were reported in the same study, the data
for the last outcome time point in each period (short or long
term) were used. Analysis was also conducted for all time points,
with the data for the last outcome time point in each study
included.
The primary outcome was a change in overall psychotic
symptoms as measured by the Brief Psychiatric Rating Scale
(BPRS) or the Positive and Negative Syndrome Scale
(PANSS).16,17 Where standard deviations for change in psychotic
symptoms were not reported, the change score was calculated
from baseline and end-point scores and the standard deviation
imputed.18 Secondary outcomes included changes in positive
and negative symptom scales. For changes in positive symptoms
we used scores from the Scale for the Assessment of Positive
Symptoms (SAPS),19 and the positive subscales of the PANSS
and BPRS. For changes in negative symptoms, scores for the
386
PANSS negative symptom subscale and the Scale for the
Assessment of Negative Symptoms (SANS) were used.20 We also
collected data on response, based on the definition by Kane et al
of a 20% decrease in BPRS total score plus either a post-treatment
Clinical Global Inventory (CGI) Scale rating of mildly ill (53) or
a post-treatment BPRS score of 35 or less.2 We have noted where
studies used other response criteria. Finally, we collected data on
those leaving the study and the following adverse drug reactions:
sialorrhoea, tachycardia, seizures, fever, dizziness, sedation,
constipation, nausea or vomiting, insomnia, dry mouth,
hypotension, headache and weight gain.
Study quality
We assessed the quality of included studies using the following
criteria adapted from Cochrane Collaboration guidelines:
(a) adequate generation of allocation sequence;
(b) masking of allocation to conditions to participant and/or
assessor;
(c) adequate random sequence generation;
(d) pre-specified primary outcome measures;
(e) appropriate reporting on missing data;
(f) use of intention-to-treat analysis;
(g) other sources of potential bias including pharmaceutical
company funding.
Statistical analysis
We used Review Manager version 5.3 for Mac for the meta-
analyses and Comprehensive Meta-Analysis version 3.3 for the
meta-regression. We calculated the standardised mean difference
(SMD) for continuous data that used different scales. We
reported the risk ratio (RR) for any dichotomous outcome. Where
possible, intention-to-treat analyses were used. We conducted
sensitivity analyses for the effect of dosages, use of first- or
second-generation control medications, pharmaceutical company
sponsorship and community or hospital study settings. We used
meta-regression to assess the effect of baseline psychosis score as
a continuous variable.21 We assessed heterogeneity using the I 2
statistic, a measure that does not depend on the number of studies
in the meta-analysis and hence has greater power to detect hetero-
geneity when the number of studies is small; it provides an
estimate of the percentage of variability due to heterogeneity
rather than chance alone. An estimate of 50% or greater indicates
possible heterogeneity, and scores of 75100% indicate considerable
heterogeneity. The I 2 estimate is calculated using the chi-squared
statistic (Q) and its degrees of freedom.18 The random effects
model was used for all the analyses as we could not definitely
exclude between-study variation even in the absence of statistical
heterogeneity, given the range of medications under review. We
tested for publication bias using funnel plot asymmetry where
low P values suggest publication bias.18
Results
A total of 2589 articles were identified in the initial search of the
electronic databases. Of these, 2402 were excluded at the title and
abstract level and a further 167 were excluded after review of the
full text (see online Fig. DS1). One additional paper was included
after a review of reference lists of key articles. Twenty-one papers
were included for the review. The sum of enrolled patients was
1131 commenced on clozapine and 1233 on control medications
with 801 clozapine and 799 control participants at final follow-up.

Study characteristics
Study quality was fair (see online Table DS1). Seventeen papers
reported adequate allocation concealment, 18 were double-blind
and 3 were blinded only to assessor. Adequate random sequence
generation was reported in 18 papers. All papers reported the
primary outcome measures and adequately described missing
data. Seventeen papers reported intention-to-treat data. Three
papers did not provide any information about the funding source.
Eleven papers reported financial support from a pharmaceutical
company (8 were funded by the manufacturer of the control
medication, 2 by the manufacturer of clozapine and 1 by a
manufacturer of both clozapine and the control medication).
Among long-term studies, all pharmaceutical funding was from
the manufacturers of control medications.
Twenty-one papers were included in the meta-analysis.2,2241
They provided data on comparisons with 25 control groups,
hereafter referred to as studies. In papers where clozapine was
compared with more than one control medication, the number
of participants in the clozapine group was divided by the number
of control medications, rounded down to the nearest integer,
and used as the number of clozapine participants in analyses
comparing clozapine with each control medication. This was done
to avoid double-counting the individuals taking clozapine. Papers
were published between 1988 and 2009. Studies reported data at
time points ranging from 6 weeks to 78 weeks. Seventeen papers
reported data on 21 short-term studies. Nine papers reported data
on 13 long-term studies.
Six papers had strict adherence to our definition of treatment-
refractory schizophrenia. Five papers included participants with
only one failed trial of an antipsychotic, three papers had trial
durations less than 6 weeks or did not report on trial duration,
and nine papers reported that previous antipsychotic trials had
a chlorpromazine-equivalent dose of less than 600 mg or did not
report a dose. Eleven papers included some participants who
had left previous antipsychotic trials owing to treatment intolerance
rather than specifically treatment failure (online Table DS1).
Control medications comprised first-generation antipsychotics
including chlorpromazine and haloperidol and second-generation
antipsychotics including olanzapine, risperidone, quetiapine and
ziprasidone (Table DS1). There was no statistically significant
difference between clozapine and control groups in terms of age
or baseline psychosis score in any of the included studies.
Psychotic symptoms
Twenty short-term studies had usable data for change in total
psychotic symptoms for 604 people given clozapine and 708
people given a control medication. The standardised mean
difference (SMD) favoured clozapine (Fig. 1). Eleven long-term
studies had usable data with 368 people given clozapine and 451
people given a control medication. In contrast to the short-term
results, there was no statistically significant difference in SMD
between the groups (Fig. 1). However, clozapine was associated
with a greater improvement in score when all time frames were
combined, using the last reported time point in each study
(SMD = 70.29, 95% CI 70.49 to 70.09, P50.005; 24 studies,
n = 1858). Clozapine had a greater effect on positive symptoms
with statistically superior outcomes at all time points (Fig. 2)
compared with negative symptoms where benefits were only seen
in the short term (Fig. 3).
Sensitivity analyses
In studies that were restricted to in-patients or had not received
pharmaceutical company funding, clozapine had significantly
Clozapine meta-analysis
better outcomes in both the short and long term. For instance,
the SMD for long-term studies without pharmaceutical company
funding was 70.67 (95% CI 71.15 to 70.19, P = 0.006; 4 studies,
n = 142). We also investigated the effect of dosing. Overall,
patients on clozapine were given 228 mg less of chlorpromazine
equivalents per day than controls (95% CI 188 to 267,
P50.0001) and 9.8 mg less of olanzapine equivalents per day
(95% CI 8.4 to 11.1, P50.0001). When studies were ranked on
degree of equivalence of dose and the half with the greatest
difference excluded, the SMD more strongly favoured clozapine
in each time frame, becoming significantly superior in the long
term (SMD = 70.42, 95% CI 70.85 to 70.01, P = 0.05; 6 studies,
n = 201). When studies that had included any treatment-intolerant
participants were excluded, there was no difference in the overall
results, with clozapine showing superior outcomes in the short
term but not in the long term. Similarly, a sensitivity analysis
on whether the study was single-blinded made no difference to
the results.
There was no difference in results for studies that used first- or
second-generation comparator antipsychotics, with clozapine
showing superior outcomes in the short term but not in the
long term. We examined clozapine against specific comparator
medications for which there were two or more studies in each
time frame. Clozapine was significantly superior to olanzapine,
haloperidol and chlorpromazine in the short term; however, there
was no significant difference between clozapine and risperidone in
the short or long term, nor against olanzapine in the long term.
Three short-term studies looked specifically at people under
age 18 years.29,30,37 When only these studies were included, the
SMD more strongly favoured clozapine, whereas when the studies
specifically looking at children and adolescents were excluded, the
SMD less strongly favoured clozapine. Finally, we explored the
effect of baseline mean psychosis score using meta-regression:
higher scores predicted greater response for clozapine in the
long term (regression coefficient 0.03, s.e. = 0.01, P = 0.0034,
t2 = 0.0008) but not the short term.
Response
Six studies defined response using the criteria outlined by Kane
et al of a greater than 20% reduction in BPRS from baseline in
the presence of a post-treatment CGI Scale score of 3 or less or
a BPRS of 35 or less (online Table DS1).2,22,23,27,37,38 A further five
studies defined response as an improvement on the BPRS or
PANSS of greater than 20%.24,28,35,36,40 A final study defined
response as a greater than 30% reduction in the BPRS plus a
post-treatment CGI score of 2 or less.30 Eight short-term studies
had usable data for 598 participants taking clozapine and 620
control group participants. People taking clozapine were significantly
more likely to respond in the short term (risk ratio (RR) = 1.17,
95% CI 1.07 to 2.73, P = 0.03; 8 studies, n = 1218). The absolute
risk reduction was 12.48% (95% CI 7.52 to 17.43). Based on this
response rate, the number needed to treat was 9. Five long-term
studies had usable data for 479 participants in the clozapine group
and 489 in the control group, with no significant difference
between clozapine and control. For all time frames combined,
results just failed to reach statistical significance (RR = 1.31, 95%
CI 0.98 to 1.70, P = 0.07; 11 studies, n = 1692).
Sensitivity analyses
There were insufficient studies to do meaningful sensitivity
analyses for the short- and long-term periods separately. Sensitivity
analyses are limited to the data for all time frames combined.
Sensitivity analysis on strictness of criteria of response or
387
Siskind et al

SMD SMD
Study or subgroup Weight (%) IV, random, 95% CI IV, random, 95% CI

Short term
Azorin et al (2001)22 7.5 70.41 (70.66, 70.16)
Bondolfi et al (1998)24 6.4 0.18 (70.24, 0.61)
Buchanan et al (1998)25 6.2 70.05 (70.50, 0.40)
Cao et al (2003)26 5.8 0.08 (70.43, 0.58)
Hong et al (1997)27 4.7 70.92 (71.59, 70.25)
Kane et al (1988)2 7.5 71.10 (71.36, 70.84)
Kane et al (2001)28 6.0 70.67 (71.15,70.19)
Kumra et al (1996)29 3.4 71.08 (72.01, 70.15)
Kumra et al (2008)30 4.9 70.32 (70.96, 0.31)
McEvoy et al (2006) Olanzapine31 4.4 70.53 (71.25, 0.19)
McEvoy et al (2006) Quetiapine31 4.1 70.71 (71.50, 0.07)
McEvoy et al (2006) Risperidone31 4.0 70.98 (71.77, 70.19)
Meltzer et al (2008)32 4.5 70.10 (70.80, 0.61)
Moresco et al (2004)33 2.7 70.85 (71.94, 0.24)
Shaw et al (2006)37 3.9 70.73 (71.54, 0.09)
Volavka et al (2002) Haloperidol39 4.9 70.35 (70.98, 0.29)
Volavka et al (2002) Olanzapine39 5.0 70.12 (70.74, 0.51)
Volavka et al (2002) Risperidone39 5.0 70.25 (70.88, 0.37)
Wahlbeck et al (2000)40 3.2 0.69 (70.26, 1.65)
Wang et al (2002)41 6.1 0.04 (70.43, 0.51)
Subtotal (95% CI) 100.0 70.39 (70.61, 70.17)
Heterogeneity. t2 = 0.15; w2 = 59.81, d.f. = 19 (P50.00001); I 2 = 68%
Test for overall effect: Z = 3.50 (P = 0.0005)

Long term
Bitter et al (2004)23 15.1 70.01 (70.34, 0.32)
McEvoy et al (2006) Olanzapine31 4.2 70.67 (71.55, 0.22)
McEvoy et al (2006) Quetiapine31 4.7 70.87 (71.70, 70.03)
McEvoy et al (2006) Risperidone31 4.3 71.29 (72.17, 70.41)
Meltzer et al (2008)32 4.9 0.01 (70.80, 0.82)
Naber et al (2005)34 13.4 0.08 (70.30, 0.46)
Sacchetti et al (2009)36 15.3 0.04 (70.29) 0.36)
Tollefson et al (2001)38 16.5 0.14 (70.15, 0.44)
Volavka et al (2002) Haloperidol39 7.2 0.12 (70.51, 0.75)
Volavka et al (2002) Olanzapine39 7.3 70.18 (70.81, 0.44)
Volavka et al (2002) Risperidone39 7.1 70.25 (70.89, 0.38)
Subtotal (95% CI) 100.0 70.11 (70.31, 0.09)
Heterogeneity. t2 = 0.04; w2 = 16.47, d.f. = 10 (P50.09); I 2 = 39%
Test for overall effect: Z = 1.06 (P = 0.29)

70.2 71 0 1 2
Favours clozapine Favours control

Fig. 1 Change in total psychotic symptoms. SMD, standardised mean difference.

equivalence of dose did not alter the absence of statistically
significant difference between the groups. When only studies using
first-generation antipsychotics as the comparator were included,
clozapine was statistically significantly more likely to lead to a
response (RR = 1.77, 95% CI 1.19 to 2.64, P = 0.008; 4 studies,
n = 164). There was no statistically significant difference between
the groups when only second-generation antipsychotics were
included. When studies with pharmaceutical funding were
excluded, response statistically significantly favoured clozapine
(RR = 1.68, 95% CI 1.20 to 2.35, P = 0.002; 6 studies, n = 208).
Study completion
The number of participants at the completion of the study
time frame was compared with the number of participants
commencing the studies. For both the short- and long-term
analyses there was no statistically significant difference between
clozapine and control antipsychotics on study completion.
Adverse drug reactions
Meta-analyses on adverse reactions were conducted for any results
reported by two or more studies. The data for the last time-point
in each study was used. It was not feasible to separate adverse
reactions into time groups as only two papers reported long-term
data. Participants taking clozapine reported significantly greater
rates of sialorrhoea, tachycardia, seizures, fever, dizziness,
sedation, constipation, and nausea and vomiting (Table 1). The
number needed to harm ranged from 4 for sialorrhoea to 19 for
nausea and vomiting. Participants taking clozapine reported
significantly lower rates of insomnia and dry mouth. There was
no significant difference for hypotension, headache or weight gain.
Publication bias
There were sufficient short-term studies to test for publication
bias for the primary outcome of change in total psychosis scores.

388
 Clozapine meta-analysis

SMD SMD
Study or subgroup Weight (%) IV, random, 95% CI IV, random, 95% CI

Short term
Azorin et al (2001)22 26.3 70.30 (70.54, 70.05)
Bondolfi et al (1998)24 14.5 0.17 (70.25, 0.60)
McEvoy et al (2006) Olanzapine31 6.4 70.43 (71.15, 0.29)
McEvoy et al (2006) Quetiapine31 5.6 70.65 (71.43, 0.13)
McEvoy et al (2006) Risperidone31 5.5 70.87 (71.66, 70.09)
Meltzer et al (2008)32 6.5 70.37 (71.09, 0.34)
Rosenheck et al (1997)35 30.1 70.15 (70.35, 0.06)
Shaw et al (2006)37 5.1 70.77 (71.59, 0.05)
Subtotal (95% CI) 100.0 70.27 (70.47, 70.08)
Heterogeneity. t2 = 0.02; w2 = 10.29, d.f. = 7 (P = 0.017); I 2 = 32%
Test for overall effect: Z = 2.73 (P = 0.006)

Long term
Kane et al (2001)28 5.9 70.61 (71.35, 0.12)
McEvoy et al (2006) Olanzapine31 4.3 70.36 (71.22, 0.51)
McEvoy et al (2006) Quetiapine31 3.5 70.85 (71.81, 0.11)
McEvoy et al (2006) Risperidone31 3.0 70.80 (71.84, 0.25)
Meltzer et al (2008)32 4.9 0.07 (70.74, 0.88)
Rosenheck et al (1997)35 48.6 70.23 (70.49, 0.03)
Tollefson et al (2001)38 29.9 70.13 (70.46, 0.19)
Subtotal (95% CI) 100.0 70.25 (70.43, 70.07)
Heterogeneity. t2 = 0.00; w2 = 4.67, d.f. = 6 (P = 0.59); I 2 = 0%
Test for overall effect: Z = 2.77 (P = 0.006)

70.2 71 0 1 2
Favours clozapine Favours control

Fig. 2 Change in positive symptoms. SMD, standardised mean difference.

SMD SMD
Study or subgroup Weight (%) IV, random, 95% CI IV, random, 95% CI

Short term
Azorin et al (2001)22 31.0 70.24 (70.49, 0.00)
Bondolfi et al (1998)24 11.6 70.02 (70.44, 0.41)
McEvoy et al (2006) Olanzapine31 4.2 70.41 (71.13, 0.30)
McEvoy et al (2006) Quetiapine31 3.7 70.45 (71.22, 0.32)
McEvoy et al (2006) Risperidone31 3.5 70.95 (71.73, 70.16)
Meltzer et al (2008)32 4.3 0.22 (70.49, 0.93)
Rosenheck et al (1997)35 41.6 70.28 (70.48, 70.07)
Subtotal (95% CI) 100.0 70.25 (70.40, 70.10)
Heterogeneity. t2 = 0.000; w2 = 6.38, d.f. = 6 (P = 0.381); I 2 = 6%
Test for overall effect: Z = 3.3 (P = 0.00091)

Long term
Bitter et al (2004)23 17.7 70.02 (70.35, 0.31)
McEvoy et al (2006) Olanzapine31 6.6 70.96 (71.87, 70.04)
McEvoy et al (2006) Quetiapine31 6.3 70.65 (71.59, 0.29)
McEvoy et al (2006) Risperidone31 5.4 70.88 (71.93, 0.17)
Meltzer et al (2008)32 7.7 0.38 (70.44, 1.20)
Rosenheck et al (1997)35 19.7 70.34 (70.60, 70.09)
Sacchetti et al (2009)36 17.9 0.23 (70.10, 0.55)
Tollefson et al (2001)38 18.7 0.21 (70.09, 0.50)
Subtotal (95% CI) 100.0 70.11 (70.39, 0.16)
Heterogeneity. t2 = 0.08; w2 = 19.55, d.f. = 6 (P = 0.007); I 2 = 64%
Test for overall effect: Z = 0.81 (P = 0.42)

70.2 71 0 1 2
Favours clozapine Favours control

Fig. 3 Change in negative symptoms. SMD, standardised mean difference.

389
Siskind et al

Table 1 Adverse drug reactions Pharmaceutical industry funding is a known source of systematic
bias either through suppression of non-favourable results or
Clozapine/
Number control
inappropriate comparator medications.42,43
of participants Sensitivity analyses for first- or second-generation comparator
studies n/n RR (95% CI) NNT/NNH antipsychotic group did not appear to affect change in psychotic
symptoms. With the exception of risperidone, clozapine was
Sialorrhoea 10 577/592 4.53 (2.508.19)*** NNH 4
superior in the short term to individual second-generation anti-
Seizures 3 167/164 3.84 (1.2112.17)* NNH 17
psychotics but failed to reach statistical significance in long-term
Tachycardia 7 431/444 3.31 (1.387.93)** NNH 7
studies. However, this finding in the long term may be con-
Fever 3 352/366 2.95 (1.455.98)** NNH 19
founded by pharmaceutical industry funding of more recent
Dizziness 7 536/549 2.13 (1.313.44)** NNH 11 comparisons with second-generation antipsychotics. When the five
Sedation 11 620/635 1.84 (1.282.68)*** NNH 7 second-generation antipsychotic studies without pharmaceutical
Constipation 7 473/489 1.63 (1.032.56)* NNH 12 funding were examined, there was a trend favouring clozapine,
Nausea/vomiting 8 557/568 1.54 (1.002.39)* NNH 19 but it failed to reach statistical significance. When dosage
Insomnia 6 370/372 0.48 (0.250.90)*** NNT 13 equivalents were compared, people taking clozapine were
Dry mouth 4 293/309 0.36 (0.140.90)*** NNT 7 receiving significantly lower doses of medication. We attempted
Hypotension 3 283/295 0.91 (0.213.94) to validate this discrepancy using two different dosage equivalence
Headache 4 425/439 0.95 (0.581.56) formulas, and both demonstrated that clozapine doses were
Weight gain 4 184/182 1.37 (0.573.30) significantly lower than those of control group medication. It is
therefore possible that lower clozapine doses may have biased
NNH, number needed to harm; NNT, number needed to treat; RR, risk ratio.
*P50.05, **P50.01, ***P50.001. the data against clozapine, although this is difficult to validate
in the absence of reported serum clozapine levels in most of the
included studies. Levels were reported in only three studies and
The Eggers regression asymmetry test did not suggest publication ranged from 281 ng/ml to 715 ng/ml;24,30,37 clozapine is therapeutic
bias (intercept 70.81, 90% CI 72.11 to 0.49, P = 0.420). at levels above 350 ng/ml.44,45 Reporting of serum clozapine levels
in future RCTs would assist in confirming whether therapeutic
doses were used.
Discussion

This study is the first systematic review and meta-analysis to look
specifically at the pharmacotherapy of treatment-refractory
schizophrenia with clozapine compared with all antipsychotics,
not solely first-generation agents. We were able to include 21
studies with 2364 participants. These included 14 papers
(n = 1379) published since the data collection periods of the
previous meta-analysis,10 all of which used a second-generation
antipsychotic as a comparator. We also used a tighter definition
of treatment-refractory schizophrenia than previous meta-
analyses, based on the criteria described by Kane et al.2 We found
that clozapine was superior to other antipsychotics in reducing
positive psychotic symptoms in both the short and long term
for people with treatment-refractory disorder. In contrast,
clozapine was not superior for negative symptoms in the long
term although it was in the short term. This, in turn, may explain
the lack of any difference in both total psychotic symptoms and
response in the long term. It is unclear why the long-term
advantages of clozapine are restricted to positive symptoms. Of
the two previous meta-analyses of the effect of study duration
on outcome, one reported that clozapine was superior in long-
term studies,15 whereas the other reported that clozapine was
superior in short-term studies.10 Clozapine was particularly
effective for more severe baseline symptoms.
The source of funding did appear to have an effect on our
results. Studies without pharmaceutical industry funding favoured
clozapine more strongly and were statistically significant for all
time frames, whereas those with such support favoured the
comparator medication. This is in contrast to earlier findings that
pharmaceutical industry funding either did not alter rates of
improvement,15 or actually increased the likelihood of a study
favouring clozapine.10 One explanation is that the earlier meta-
analyses included comparisons with first-generation antipsychotics
that were funded by manufacturers of clozapine. In contrast, all
but two of the additional 14 papers included in this review were
comparisons with second-generation antipsychotics that were
funded by the manufacturers of the second-generation agents.
Study limitations
There were several limitations of this study. Many of our results
showed heterogeneity. Although we attempted to explore this
further with sensitivity analyses and meta-regression as well as
using a random effects model throughout to incorporate hetero-
geneity into our analysis, our results should still be treated with
caution. As noted above, there are potential biases associated with
pharmaceutical industry funding, class of control antipsychotic
and comparative doses of clozapine and control medications.
We attempted to address these by conducting sensitivity analyses
and meta-regression. It is important to note the difficulty of
masking in studies where clozapine is a comparator, given the
significant adverse drug reactions associated with clozapine. It is
possible that this may have systematically biased the overall
results. Several studies included participants who had been
intolerant to previous antipsychotic trials, as opposed to a strict
definition of two or more failed adequate trials. A sensitivity
analysis of studies with strict inclusion criteria did not alter the
results. We were unable to report on relapse, as the included
studies did not provide usable data on this variable. Although
we attempted to locate unpublished findings, it is possible that
there are unpublished data that we were unable to include.
Clinical implications
Our results suggest that clozapine should remain the treatment of
choice for refractory schizophrenia, at least in the short term.
Clozapine demonstrated superiority for positive symptoms across
all time frames. Given the challenges associated with treating
people with refractory disorder, our finding of a number needed
to treat of 9 is moderately good.46 However, this must be balanced
against numbers needed to harm that ranged from 4 for
sialorrhoea to 19 for fever. In addition, if there is no meaningful
improvement of symptoms or function at 6 months, our findings
suggest clozapine should be stopped and consideration given to an
antipsychotic with a more favourable adverse reaction profile.

390

Pharmacological treatment should always be provided in concert
with evidence-based psychosocial interventions.47
Dan Siskind, MBBS, PhD, Metro South Addiction and Mental Health Services,
Brisbane, and University of Queensland School of Medicine, Brisbane; Lara
McCartney, MBBS, Royal Melbourne Hospital, Melbourne; Romi Goldschlager,
MBBS, St Vincents Hospital, Melbourne; Steve Kisely, MD, PhD, Metro South
Addiction and Mental Health Services, Brisbane, University of Queensland School
of Medicine, Brisbane, and Griffith Health Institute, Brisbane, Australia
Correspondence: Dan Siskind, MIRT, 519 Kessels Road, MacGregor,
Queensland 4109, Australia. Email: d.siskind@uq.edu.au
First received 13 Jul 2015, final revision 13 Dec 2015, accepted 27 Feb 2016
Funding
D.S. is funded in part through an NHMRC ECF, APP1111136, 20162019.
Acknowledgements
We would like to thank Evelyn Ma, Dagmar Hlincikova and Oliver Freudenreich for their
assistance in translating articles.
References
1 Idanpaan-Heikkila J, Alhava E, Olkinuora M, Palva IP. Agranulocytosis during
treatment with chlozapine. Eur J Clin Pharmacol 1977; 11: 1938.
2 Kane J, Honigfeld G, Singer J, Meltzer H. Clozapine for the treatment-resistant
schizophrenic. A double-blind comparison with chlorpromazine. Arch Gen
Psychiatry 1988; 45: 78996.
3 Leucht S, Cipriani A, Spineli L, Mavridis D, Orey D, Richter F, et al.
Comparative efficacy and tolerability of 15 antipsychotic drugs in
schizophrenia: a multiple-treatments meta-analysis. Lancet 2013; 382:
95162.
4 Lehman AF, Lieberman JA, Dixon LB, McGlashan TH, Miller AL, Perkins DO,
et al. Practice guideline for the treatment of patients with schizophrenia,
second edition. Am J Psychiatry 2004; 161 (suppl 2): 156.
5 National Institute for Health and Care Excellence. Psychosis and
Schizophrenia in Adults: Prevention and Management. CG 178. NICE, 2014.
6 Royal Australian and New Zealand College of Psychiatrists Clinical Practice
Guidelines Team for the Treatment of Schizophrenia and Related Disorders.
Royal Australian and New Zealand College of Psychiatrists clinical practice
guidelines for the treatment of schizophrenia and related disorders. Aust NZ
J Psychiatry 2005; 39: 130.
7 Warnez S, Alessi-Severini S. Clozapine: a review of clinical practice guidelines
and prescribing trends. BMC Psychiatry 2014; 14: 102.
8 Forrester T, Siskind D, Winckel K, Wheeler A, Hollingworth S. Increasing
clozapine dispensing trends in Queensland, Australia 20042013.
Pharmacopsychiatry 2015; 48: 1649.
9 Howes OD, Vergunst F, Gee S, McGuire P, Kapur S, Taylor D. Adherence to
treatment guidelines in clinical practice: study of antipsychotic treatment
prior to clozapine initiation. Br J Psychiatry 2012; 201: 4815.
10 Moncrieff J. Clozapine v. conventional antipsychotic drugs for treatment-
resistant schizophrenia: a re-examination. Br J Psychiatry 2003; 183: 1616.
11 Booth A, Clarke M, Dooley G, Ghersi D, Moher D, Petticrew M, et al. The nuts
and bolts of PROSPERO: an international prospective register of systematic
reviews. Syst Rev 2012; 1: 2.
12 Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA group (2009)
preferred reporting items for systematic reviews and meta-analyses: the
PRISMA statement. PLoS Med 2009; 6: e1000097.
13 Andreasen NC, Pressler M, Nopoulos P, Miller D, Ho BC. Antipsychotic dose
equivalents and dose-years: a standardized method for comparing exposure
to different drugs. Biol Psychiatry 2010; 67: 25562.
14 Leucht S, Samara M, Heres S, Patel MX, Furukawa T, Cipriani A, et al. Dose
equivalents for second-generation antipsychotic drugs: the classical mean
dose method. Schizophr Bull 2015; 41: 1397402.
15 Wahlbeck K, Cheine M, Essali A, Adams C. Evidence of clozapines
effectiveness in schizophrenia: a systematic review and meta-analysis of
randomized trials. Am J Psychiatry 1999; 156: 9909.
16 Overall JE, Gorham DR. The brief psychiatric rating scale. Psychol Rep 1962;
10: 799812.
17 Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale
(PANSS) for schizophrenia. Schizophr Bull 1987; 13: 26176.
Clozapine meta-analysis
18 Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of
Interventions Version 5.1.0. Cochrane Collaboration, 2011.
19. Andreasen NC. The Scale for the Assessment of Positive Symptoms (SAPS).
University of Iowa, 1984.
20 Andreasen NC. The Scale for the Assessment of Negative Symptoms (SANS).
University of Iowa, 1983.
21 Leucht S, Rothe P, Davis JM, Engel RR. Equipercentile linking of the BPRS and
the PANSS. Eur Neuropsychopharmacol 2013; 23: 9569.
22 Azorin JM, Spiegel R, Remington G, Vanelle JM, Pere JJ, Giguere M, et al. A
double-blind comparative study of clozapine and risperidone in the
management of severe chronic schizophrenia. Am J Psychiatry 2001; 158:
130513.
23 Bitter I, Dossenbach MR, Brook S, Feldman PD, Metcalfe S, Gagiano CA, et al.
Olanzapine versus clozapine in treatment-resistant or treatment-intolerant
schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 2004; 28:
17380.
24 Bondolfi G, Dufour H, Patris M, May JP, Billeter U, Eap CB, et al. Risperidone
versus clozapine in treatment-resistant chronic schizophrenia: a randomized
double-blind study. The Risperidone Study Group. Am J Psychiatry 1998; 155:
499504.
25 Buchanan RW, Breier A, Kirkpatrick B, Ball P, Carpenter WT. Positive and
negative symptom response to clozapine in schizophrenic patients with and
without the deficit syndrome. Am J Psychiatry 1998; 155: 75160.
26 Cao HJ, You HF, Fan FL, Zhang J. The control study of risperidone and
clozapine for the treatment-resistant schizophrenia. J Chinese Med Res 2003;
4: 316-9.
27 Hong CJ, Chen JY, Chiu HJ, Sim CB. A double-blind comparative study of
clozapine versus chlorpromazine on Chinese patients with treatment-
refractory schizophrenia. Int Clin Psychopharmacol 1997; 12: 12330.
28 Kane JM, Marder SR, Schooler NR, Wirshing WC, Umbricht D, Baker RW, et al.
Clozapine and haloperidol in moderately refractory schizophrenia: a 6-month
randomized and double-blind comparison. Arch Gen Psychiatry 2001; 58:
96572.
29 Kumra S, Frazier JA, Jacobsen LK, McKenna K, Gordon CT, Lenane MC, et al.
Childhood-onset schizophrenia. A double-blind clozapinehaloperidol
comparison. Arch Gen Psychiatry 1996; 53: 10907.
30 Kumra S, Kranzler H, Gerbino-Rosen G, Kester HM, Thomas C, Kafantaris V,
et al. Clozapine and high-dose olanzapine in refractory early-onset
schizophrenia: a 12-week randomized and double-blind comparison. Biol
Psychiatry 2008; 63: 524-9.
31 McEvoy JP, Lieberman JA, Stroup TS, Davis SM, Meltzer HY, Rosenheck RA,
et al. Effectiveness of clozapine versus olanzapine, quetiapine, and
risperidone in patients with chronic schizophrenia who did not respond
to prior atypical antipsychotic treatment. Am J Psychiatry 2006; 163:
60010.
32 Meltzer HY, Bobo WV, Roy A, Jayathilake K, Chen Y, Ertugrul A, et al.
A randomized, double-blind comparison of clozapine and high-dose
olanzapine in treatment-resistant patients with schizophrenia. J Clin
Psychiatry 2008; 69: 27485.
33 Moresco RM, Cavallaro R, Messa C, Bravi D, Gobbo C, Galli L, et al. Cerebral
D2 and 5-HT2 receptor occupancy in schizophrenic patients treated with
olanzapine or clozapine. J Psychopharmacol 2004; 18: 35565.
34 Naber D, Riedel M, Klimke A, Vorbach EU, Lambert M, Kuhn KU, et al.
Randomized double blind comparison of olanzapine vs. clozapine on
subjective well-being and clinical outcome in patients with schizophrenia.
Acta Psychiatr Scand 2005; 111: 10615.
35 Rosenheck R, Cramer J, Xu W, Thomas J, Henderson W, Frisman L, et al.
A comparison of clozapine and haloperidol in hospitalized patients with
refractory schizophrenia. N Engl J Med 1997; 337: 80915.
36 Sacchetti E, Galluzzo A, Valsecchi P, Romeo F, Gorini B, Warrington L,
et al. Ziprasidone vs clozapine in schizophrenia patients refractory to
multiple antipsychotic treatments: the MOZART study. Schizophr Res 2009;
110: 809.
37 Shaw P, Sporn A, Gogtay N, Overman GP, Greenstein D, Gochman P, et al.
Childhood-onset schizophrenia: a double-blind, randomized clozapine-
olanzapine comparison. Arch Gen Psychiatry 2006; 63: 72130.
38 Tollefson GD, Birkett MA, Kiesler GM, Wood AJ. Double-blind comparison of
olanzapine versus clozapine in schizophrenic patients clinically eligible for
treatment with clozapine. Biol Psychiatry 2001; 49: 5263.
39 Volavka J, Czobor P, Sheitman B, Lindenmayer JP, Citrome L, McEvoy JP,
et al. Clozapine, olanzapine, risperidone, and haloperidol in the treatment
of patients with chronic schizophrenia and schizoaffective disorder.
Am J Psychiatry 2002; 159: 255-62
40 Wahlbeck K, Cheine M, Tuisku K, Ahokas A, Joffe G, Rimon R. Risperidone
versus clozapine in treatment-resistant schizophrenia: a randomized pilot
study. Prog Neuropsychopharmacol Biol Psychiatry 2000; 24: 91122.
391
Siskind et al

41 Wang RZ, Geng YY, Pan DH, Zhang SJ. A comparative trial of efficacy of
risperidone vs clozapine in treatment of refractory schizophrenia. Chinese
J Pharmacoepidemiol 2002; 5: 2301.
42 Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective
publication of antidepressant trials and its influence on apparent efficacy.
N Engl J Med 2008; 358: 25260.
43 Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry
sponsorship and research outcome and quality: systematic review. BMJ
2003; 326: 116770.
44 Stark A, Scott J. A review of the use of clozapine levels to guide treatment
and determine cause of death. Aust NZ J Psychiatry 2012; 46: 81625.
45 Potkin SG, Bera R, Gulasekaram B, Costa J, Hayes S, Jin Y, et al. Plasma
clozapine concentrations predict clinical response in treatment-resistant
schizophrenia. J Clin Psychiatry 1994; 55 (suppl B): 1336.
46 Citrome L, Ketter TA. When does a difference make a difference?
Interpretation of number needed to treat, number needed to harm, and
likelihood to be helped or harmed. Int J Clin Pract 2013; 67: 40711.
47 Dixon LB, Dickerson F, Bellack AS, Bennett M, Dickinson D, Goldberg RW,
et al. The 2009 schizophrenia PORT psychosocial treatment
recommendations and summary statements. Schizophr Bull 2010;
36: 4870.

Suicide Junkie
poems
by Jo McFarlane
doctors
The last overdose I took
was precipitated by the fear of abandonment.
I had misinterpreted my key workers kindness
as a sign she was soon to be leaving,
I wasnt trying to make her stay;
I just couldnt imagine how Id cope without her
so death seemed the only viable option.
When I was younger
I used to fantasise about being murdered.
It seemed the obvious solution
to my impossible predicament:
Life too unbearable to go on
but I wouldnt have to face the wrath of hell
because the fatal knife plunged into me
would be the doing of anothers hand.
So Id tank up on the cheapest and most potent booze
then stagger round the streets all night,
searching for a friendly psychopath
to put me out of my misery.
Some nights I slept on a dried patch of dogs piss,
soaking up the fumes with a vampires
lust for the macabre.
There was the odd attempted hanging
where I proved myself a parody
of beefed-up incompetence.
I even tried to drown myself by swimming out to sea.
The life crew that picked me up
were unaware of the irony in my thanks
When I jumped off Salisbury Crags
some unseen force intervened and I came off lightly
with a thousand bruises and a broken bone.
Staring from the Forth Road Bridge into the black abyss,
I should have known
the fence would prove too much for my dyspraxia.
Each botched attempt to silence the scream
another notch upon the damning pages of my notes.
Lurching from crisis to crisis
fed the adrenalin of being mad.
I wasnt doing it for attention
but release from the distress.
Its hard to explain if youve not known the pain
of desolation in the beauty of a sunset.
Being addicted to despair is a sickness of the soul.
The healing touch of heaven, I now see,
comes not from death
but the sunset quickening my senses,
making me feel whole.
B Jo McFarlane. Reprinted with permission.
Selected by Femi Oyebode. From Stigma & Stones: Living with a Diagnosis of BPD, poems by Sally Fox & Jo McFarlane.
Through their collection Stigma & Stones, writers/performers/partners Sally Fox and Jo McFarlane seek to promote
understanding, improve treatment and reduce the stigma of living with a diagnosis of BPD.

The British Journal of Psychiatry (2016)

209, 392. doi: 10.1192/bjp.bp.116.181669

392
Data supplement to Siskind et al. Clozapine v. first- and second-generation
antipsychotics in treatment-refractory schizophrenia: systematic review and
meta-analysis. Br J Psychiatry doi: 10.1192/bjp.bp.115.177261

Fig DS1 PRISMA diagram

*Unique articles identified through literature search, n = 2589

Cochrane n = 934
EMBASE n = 688
Pubmed, n = 967

Excluded on title and abstract and duplicates

removed, n = 2402

Screening of full text for inclusion

n = 187

Excluded, n = 165
Wrong comparator group = 14
Not treatment resistant = 61
Not randomized controlled trial = 50
Nil useable data = 8
Redacted post publication = 2
Duplicated publications from single
study = 46
*Some studies fell under multiple
categories

Additional study included from search of

reference lists, n = 1

RCTs included in meta-analysis

n = 21
Table DS1 Included studies
Mean Age Number of CPZ CPZ
Primary
Setting Diagnostic TRS criteria Control (SD) for Participants equivalent equivalent Random Repor Other
Paper1 Duration2 Country Allocation911 Blinding10 Outcom 13 ITT
1416
3
tool4 deviation5 Medication6 Clozapine / Clozapine / control clozapine isation11 ting Bias1517
e12
Control7 Control (SD)8 (SD)

France & One Trial, A 37.8 (10.4) / Pharma

Azorin et al 2001 12 w H+C DSM-IV Risperidone 138 / 135 NS NS NS Double Yes Yes Yes Yes
Canada intolerance 39.5 (11.3) clozapine
Hungary &
One Trial, A Pharma
Bitter et al 2004 18 w South H DSM-IV Olanzapine 37.6 * 72 / 72 441 (102) 224 (100) NS Double Yes Yes Yes Yes
intolerance control
Africa
Bondolfi et al Switzerlan No Dose, A 36.2 (12.2) Pharma
8w C DSM-III-R Risperidone 43 / 43 NS NS NS Double Yes Yes Yes Yes
1998 d intolerance / 38.3 (12.9) control
Pharma
Buchanan et al T 41.0 (6.4) /
10 w USA C DSM-III-R No deviation Haloperidol 38 / 37 1124 (334) 474 (50) Yes Double Yes Yes Yes Yes clozapine &
1998 40.1 (7.9) control
A 36.9 (7.9) /
Cao et al 2003 12 w China H ICD 10 No deviation Risperidone 30 / 30 NS NS NS Double Yes Yes Yes Yes FNS
37.5 (8.7)

T 39.7 (8.4) /
Hong et al 1997 12 w China H DSM-IV No deviation Chlorpromzaine 21 / 19 1163 (228) 641 (155) Yes Double Yes Yes Yes Yes No
37.1 (8.7)

T Pharma
Kane et al 1988 6w USA H DSM-III No deviation Chlorpromzaine 35.7 (8.87)* 37 / 34 NS NS NS Double NS Yes Yes Yes
clozapine

41 (10) /
Kane et al 2001 29 w16 USA H+C DSM-III-R No deviation Haloperidol T
126 / 141 900 (334) 523 (171) Yes Double Yes Yes Yes No No
40 (8)

No Dose or
Kumra et al T 14.4 (3.0) /
6w USA H DSM-III-R Duration, Haloperidol 10 / 11 718 (378) 176 (145) Yes Double Yes Yes Yes Yes No
1996 13.7 (1.6)
intolerance
Kumra et al not No Dose or A 15.8 (2.2) /
12 w USA DSM-IV Olanzapine 18 / 21 716 (144) 461 (207) Yes Double Yes Yes Yes Yes No
2008 stated Duration 15.5 (2.1)
39.4 (9.9) / 43 / Clozapine
One Trial & 629 (180)
McEvoy et al Olanzapine A 44.3 (10.5) / 17 / single,
78 w17 USA C DSM-IV no Dose, 398 (97) 368 (154) NS Yes Yes Yes Yes No
2006 Risperidone A 37.1 (11.8) / 14 / control
intolerance 589 (145)
Quetiapine A 39.7 (10.4) 14 double

Meltzer et al 37.2 (9.2) / Pharma

26 w18 USA C DSM-IV No Dose Olanzapine A
21 / 19 953 (270) 680 (256) Yes Double NS Yes Yes No
2008 36.4 (11.1) control

Moresco et al A 38.3 (9.1) / Pharma

8w Italy H DSM-IV No deviation Olanzapine 12 / 11 474 (8) 359 (6) Yes Double NS Yes Yes No
2004 34.1 (7.6) control

One Trial &

A 35.2 (10.8) / Pharma
Naber et al 2005 26 w Germany H+C DSM-IV no Dose, Olanzapine 57 / 57 412 (102) 215 (82) NS Double Yes Yes Yes Yes
32.9 (10.4) control
intolerance
 Rosenheck et al 43.2 (7.7) /
52 w19 USA H+C DSM-III-R Intolerance Haloperidol T
205 / 218 NS NS NS Double Yes Yes Yes Yes No
1997 43.9 (8.3)

Sacchetti et al A 38.3 (11.2) / Pharma

18 w Italy C DSM-IV Intolerance Ziprasidone 74 / 73 450 (31) 386 (52) NS Double Yes Yes Yes No
2009 41.6 (10.2) control
<4 week
Duration
>100mg A 12.8 (2.4) /
Shaw et al 2006 8w USA H K-SADS Olanzapine 12 / 13 468 (90) 362 (106) Yes Double Yes Yes Yes Yes FNS
CPZ 11.7 (2.3)
equivalent,
intolerance
>500mg
Tollefson et al CPZ Pharma
18 w Multiple20 C DSM-IV Olanzapine A
38.6 (10.6)* 90 / 90 450 (55) 332 (101) NS Double NS Yes Yes Yes
2001 equivalent, control
intolerance
Olanzapine A 40 / 41 513 (39)
Volavka et al Pharma
14 w21 USA H DSM-IV One Trial Risperidone A 40.8 (9.2)* / 39 683 (163) 459 (158) Yes Double Yes Yes Yes Yes
2002 control
Haloperidol T / 37 837 (157)

Wahlbeck et al A 35.7 (9.8) /

10 w Finland H+C DSM-IV Intolerance Risperidone 11 / 9 673 (163) 437 (227) Yes Single Yes Yes Yes Yes No
2000 36.8 (9.8)

A 35.6 (7.5) /
Wang et al 2002 12 w China H CCMD-3 No Dose Risperidone 35 / 35 NS NS NS Single Yes Yes Yes Yes FNS
36 (7.5)

NS=not stated

1
Lead author and year of publication
2
w=weeks
3
H = Hospital, C = Community
4
DSM = Diagnostic and Statistical Manual, ICD = International Classification of Diseases, K-SADS = Schedule for Affective Disorders and Schizophrenia for
School-Age Children CCMD-3 = Chinese classification of mental disorder
5
TRS= Treatment Refractory Schizophrenia. Table lists areas in which study varied from definition of TRS criteria of failed treatment: Trial of 2
antipsychotics; Duration 6 weeks each; Dose over 600mg/day chlorpromazine (CPZ) equivalents); trials not shortened because of intolerable side effects.
6
A=Atypical antipsychotic, T=Typical antipsychotic
7
SD= Standard Deviation, * = mean age and SD not provided for both clozapine and control
8
Mean chlorpromazine equivalent dose and standard deviation of control medication and clozapine using power transformation formula from Andreason et al
(2010)
9
Adequate allocation concealment
10
Single is to assessor only
11
Adequate Random Sequence Generation
12
Primary Outcome Measures were pre-specified and reported
13
Completeness of outcome reporting
14
ITT=Intention to Treat analysis
15
Were other potential sources of bias present. Pharma=Pharmaceutical company sponsorship. FNS = funding source not specified
16
Data reported at 5, 11, 17 and 29 weeks
17
Data reported at 3 and 6 months
18
Data reported at 6 weeks and 6 months
19
Data reported at 6 weeks, 3 months, 6 months, 9 months and 12 months
20
Belgium, Denmark, Finland, France, Germany, Italy, Norway, Portugal, South Africa, Spain, Sweden, Switzerland, Great Britain, and Ireland
21
Data reported at 8 and 14 weeks
References
Azorin JM, Spiegel R, Remington G, Vanelle JM, Pere JJ, Giguere M,

Bourdeix I. A double-blind comparative study of clozapine and risperidone in

the management of severe chronic schizophrenia. Am J Psychiatry 2001,

Aug;158(8):1305-13.

Bitter I, Dossenbach MR, Brook S, Feldman PD, Metcalfe S, Gagiano CA, et

al. Olanzapine versus clozapine in treatment-resistant or treatment-intolerant

schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 2004,

Jan;28(1):173-80.

Bondolfi G, Dufour H, Patris M, May JP, Billeter U, Eap CB, Baumann P.

Risperidone versus clozapine in treatment-resistant chronic schizophrenia: A

randomized double-blind study. The risperidone study group. Am J Psychiatry

1998, Apr;155(4):499-504.

Buchanan RW, Breier A, Kirkpatrick B, Ball P, Carpenter WT. Positive and

negative symptom response to clozapine in schizophrenic patients with and

without the deficit syndrome. Am J Psychiatry 1998(6):751-60.

Cao HJ, You HF, Fan FL, Zhang J. The control study of risperidone and

clozapine for the treatment-resistant schizophrenia 2003(4):316-9

Hong CJ, Chen JY, Chiu HJ, Sim CB. A double-blind comparative study of

clozapine versus chlorpromazine on chinese patients with treatment-refractory

schizophrenia 1997(3):123-30

Kane JM, Marder SR, Schooler NR, Wirshing WC, Umbricht D, Baker RW, et

al. Clozapine and haloperidol in moderately refractory schizophrenia: A 6-

month randomized and double-blind comparison. Arch Gen Psychiatry 2001,

Oct;58(10):965-72.

Kumra S, Frazier JA, Jacobsen LK, McKenna K, Gordon CT, Lenane MC, et

al. Childhood-onset schizophrenia. A double-blind clozapine-haloperidol

comparison. Arch Gen Psychiatry 1996, Dec;53(12):1090-7.

Kumra S, Kranzler H, Gerbino-Rosen G, Kester HM, Thomas C, Kafantaris V,

et al. Clozapine and "high-dose" olanzapine in refractory early-onset

schizophrenia: A 12-week randomized and double-blind comparison

2008(5):524-9

McEvoy JP, Lieberman JA, Stroup TS, Davis SM, Meltzer HY, Rosenheck

RA, et al. Effectiveness of clozapine versus olanzapine, quetiapine, and

risperidone in patients with chronic schizophrenia who did not respond to prior

atypical antipsychotic treatment. Am J Psychiatry 2006, Apr;163(4):600-10.

Meltzer HY, Bobo WV, Roy A, Jayathilake K, Chen Y, Ertugrul A, et al. A

randomized, double-blind comparison of clozapine and high-dose olanzapine

in treatment-resistant patients with schizophrenia. Journal of Clinical

Psychiatry 2008, Feb;69 (2):274-85.

Moresco RM, Cavallaro R, Messa C, Bravi D, Gobbo C, Galli L, et al. Cerebral

D2 and 5-HT2 receptor occupancy in schizophrenic patients treated with

olanzapine or clozapine. J Psychopharmacol 2004, Sep;18(3):355-65.

Naber D, Riedel M, Klimke A, Vorbach EU, Lambert M, Kuhn KU, et al.

Randomized double blind comparison of olanzapine vs. Clozapine on

subjective well-being and clinical outcome in patients with schizophrenia. Acta

Psychiatr Scand 2005, Feb;111(2):106-15.

Rosenheck R, Cramer J, Xu W, Thomas J, Henderson W, Frisman L, et al. A

comparison of clozapine and haloperidol in hospitalized patients with

refractory schizophrenia. New England Journal of Medicine 1997, Sep 18;337

(12):809-15.

Sacchetti E, Galluzzo A, Valsecchi P, Romeo F, Gorini B, Warrington L,

INITIATE Study Group. Ziprasidone vs clozapine in schizophrenia patients

refractory to multiple antipsychotic treatments: The MOZART study. Schizophr

Res 2009, May;110(1-3):80-9.

Shaw P, Sporn A, Gogtay N, Overman GP, Greenstein D, Gochman P, et al.

Childhood-onset schizophrenia: A double-blind, randomized clozapine-

olanzapine comparison. Arch Gen Psychiatry 2006;63(7):721-30.

Tollefson GD, Birkett MA, Kiesler GM, Wood AJ. Double-blind comparison of

olanzapine versus clozapine in schizophrenic patients clinically eligible for

treatment with clozapine. Biol Psychiatry 2001, Jan 1;49(1):52-63.

Volavka J, Czobor P, Sheitman B, Lindenmayer JP, Citrome L, McEvoy JP, et

al. Clozapine, olanzapine, risperidone, and haloperidol in the treatment of

patients with chronic schizophrenia and schizoaffective disorder 2002(2):255-

62

Wahlbeck K, Cheine M, Tuisku K, Ahokas A, Joffe G, Rimon R. Risperidone

versus clozapine in treatment-resistant schizophrenia: A randomized pilot

study. Prog Neuropsychopharmacol Biol Psychiatry 2000;24(6):911-22.

Wang RZ, Geng YY, Pan DH, Zhang SJ. A comparative trial of efficacy of

risperidone vs clozapine in treatment of refractory schizophrenia. Chinese

Journal of Pharmacoepidemiology 2002(5):230-1.

Clozapine v. first- and second-generation antipsychotics in
treatment-refractory schizophrenia: systematic review and
meta-analysis
Dan Siskind, Lara McCartney, Romi Goldschlager and Steve Kisely
BJP 2016, 209:385-392.
Access the most recent version at DOI: 10.1192/bjp.bp.115.177261

Supplementary Supplementary material can be found at:

Material http://bjp.rcpsych.org/content/suppl/2016/06/29/bjp.bp.115.177261.DC1.html

References This article cites 43 articles, 10 of which you can access for free at:
http://bjp.rcpsych.org/content/209/5/385#BIBL
Reprints/ To obtain reprints or permission to reproduce material from this paper, please
permissions write to permissions@rcpsych.ac.uk

You can respond /letters/submit/bjprcpsych;209/5/385

to this article at
Downloaded http://bjp.rcpsych.org/ on January 31, 2017
from Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/site/subscriptions/